First Pass
I’ve always been head-over-heels about sports. I played three sports all through high school, was a four-year varsity athlete in college, and still play in a number of men’s leagues. I’m also a huge fan, watching and reading everything I can get my hands on. Over the last few year, I’ve become particularly interested in the analytics movement as a new lens to examine and appreciate different games. I’m using this blog as a platform to share what I’m working on and hopefully develop my own skills.
For this first piece (and probably a number afterwards), I’m digging into a dataset that StatsBomb has made public, containing individual event data from all 64 games in the 2018 World Cup and analyzing the data using R.
Today I’m going to be focusing on France, the world champions. A driving force behind their victory was Paul Pogba, their 25-year-old midfield engine. He scored France’s third goal in the 59th minute of the Final to put France up 3–1, but his primary function, and where he truly shined all tournament, was as a hub in the midfield, using his fantastic technique and vision to progress the ball up the pitch to France’s forwards.
This YouTube compilation from the final (obviously none of the video’s mine, sorry FIFA) does a good job demonstrating Pogba’s abilities as a ball-distributor. He owned the center of the park in almost every game he played, completing 244 of his 305 pass attempts over the course of the tournament.
After cleaning and combining the individual game data to create a giant dataset with all game events, I filtered it to included only passes by Pogba in this basic scatter plot (created using the ggplot2 package), which shows the distribution of locations that he passed from.
That’s definitely a lot to take in, and hard to get a pattern from. Instead, I’m going to look at this as a density graph. This is color coded (in French colors!) based on the density of passes in the vicinity. I also filtered the dataset further to only include completed passes, set limits on the scales so that only the 120x80 field was visible, and flipped the y-axis to correspond with the where the passes were actually distributed (the higher xlocation numbers correspond to the right side of the field.
Viola. That’s certainly easier to pick up on right away. Pogba operated as a right central midfielder for most of the tournament, alongside N’Golo Kante, and it’s clear from the density chart that that was definitely his central location. But where is the ball going from there? Let’s look at the same type of graph, but with the pass destinations instead of the pass origins.
As expected, Pogba was spraying passes all over the yard, though with a definite focus towards the right side. If anyone watched much of the World Cup, you almost certainly noticed Kylian Mbappé, France’s electric right winger, who scored 4 goals and terrorized defenses with his speed. Pogba linked up with Mbappé 31 times over the course of France’s 7 game tear.
I’ll dig more into Mbappé going forward, but this is a look at where those link-ups between Pogba and Mbappé occurred.
Surprise! One of the world’s most dangerous right wingers prefers to receive the ball on the right side. It’s cool to see the numbers back up the eye test like that.
This last chart, which I really like, shows the shape of Pogba’s passes relative to himself.
I cheated a little bit on this one, as I added the cross after the fact, just to make it clearer where the origin point was. He’s clearly a hub, with a solid distribution of passes all around him. I’ll compare this graph to that of others in future posts.
As the title implies, this is a first pass at all of this. I am going to continue to improve at this and would love any feedback or suggestions for what I can look at next!
